michal
commited on
Commit
•
24d957e
1
Parent(s):
d6e7632
Only train split
Browse files- biovdb_1000.py +3 -5
biovdb_1000.py
CHANGED
@@ -3,7 +3,7 @@ import datasets
|
|
3 |
|
4 |
CITATION = """
|
5 |
"""
|
6 |
-
|
7 |
CSV_URL = "https://huggingface.co/datasets/mwinn99/biovdb_1000/resolve/main/biovdb_1000.csv"
|
8 |
|
9 |
class Biovdb(datasets.GeneratorBasedBuilder):
|
@@ -23,10 +23,8 @@ class Biovdb(datasets.GeneratorBasedBuilder):
|
|
23 |
def _split_generators(self, dl_manager):
|
24 |
data_fname = dl_manager.download(CSV_URL)
|
25 |
data = pd.read_csv(data_fname)
|
26 |
-
series_ids = data['series_id'].str.split(',', expand=True).stack().unique()
|
27 |
return [
|
28 |
-
datasets.SplitGenerator(name=
|
29 |
-
for series_id in series_ids
|
30 |
]
|
31 |
|
32 |
def _generate_examples(self, data):
|
@@ -41,5 +39,5 @@ class Biovdb(datasets.GeneratorBasedBuilder):
|
|
41 |
}
|
42 |
|
43 |
if __name__ == '__main__':
|
44 |
-
ds = datasets.load_dataset("./biovdb_1000.py",
|
45 |
|
|
|
3 |
|
4 |
CITATION = """
|
5 |
"""
|
6 |
+
#TODO: There must be other way to refer to file in the repo
|
7 |
CSV_URL = "https://huggingface.co/datasets/mwinn99/biovdb_1000/resolve/main/biovdb_1000.csv"
|
8 |
|
9 |
class Biovdb(datasets.GeneratorBasedBuilder):
|
|
|
23 |
def _split_generators(self, dl_manager):
|
24 |
data_fname = dl_manager.download(CSV_URL)
|
25 |
data = pd.read_csv(data_fname)
|
|
|
26 |
return [
|
27 |
+
datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"data": data})
|
|
|
28 |
]
|
29 |
|
30 |
def _generate_examples(self, data):
|
|
|
39 |
}
|
40 |
|
41 |
if __name__ == '__main__':
|
42 |
+
ds = datasets.load_dataset("./biovdb_1000.py", split='train')
|
43 |
|